Abstract: We study the coherence of one-year continuous waveforms from the Piñon Flats Observatory Array (PY) and five dense linear arrays (JF, DW, SGB, RA, and BB from south to north) along the San Jacinto fault using the multitaper spectral analysis. The examined data include ambient noise and earthquake signals from local and teleseismic sources recorded by different types of seismometers at different locations. In the PY array, the coherence of noise data remains high in the 0.1-1.0 Hz band when the interstation distance is small (<65 m), and decreases quickly outside the dominant frequency band and with increasing station distances. The local earthquake signals exhibit high coherence up to 20 Hz, while the teleseismic waves have high coherences in the low frequency band (<0.1 Hz). The coherences show complex daily variations up to 12 Hz. The high coherence (>0.95) frequency bands in the two horizontal components exhibit seasonal variations. In addition, the coherences of horizontal components are higher in the 2-5 Hz band when compared to the vertical component, while the reverse is true in the low frequency range (<0.1 Hz). The coherence at PY also contains high anomalies in ~2-4 Hz lasting for ~16 days before the 2016 M5.2 Borrego Springs event, which might be related to anthropogenic effects. In the JF and DW arrays, coherence exhibits similar patterns in 0.1-1.0 Hz without significant daily or seasonal changes, while in the SGB and RA arrays there are no highly coherent noise signals in 0.1-1.0 Hz.
Media has not been submitted for this Presentation
Abstract: Although it is generally thought that the Earth’s outer core is well mixed due to vigorous convection, and is therefore more or less compositionally homogeneous, there are, however, increasing evidence that suggests stratification may exist within the outermost outer core due to the presence of light elements. Recent seismic studies show that top a few hundreds kilometers of outer core possess a P-wave velocity slightly lower than the PREM model, which cannot be explained by self-compression of a chemically homogeneous outer core. Most studies utilize the SmKS arrival, which is a core phase being reflected m-1 times from the lower side of the core-mantle boundary (CMB) and can be observed at epicentral distances of 120°-180°. Differential arrival times between SmKS pairs, such as S3KS and SKKS, S4KS and S3KS, are usually employed in determine the P-wave velocity structure in the top part of the outer core since these pairs have very similar traveling ray paths in the mantle. Since there is a π/2 phase shift between two consecutive SmKS arrivals due to the internal caustic surface for underside reflection, measuring the differential times between the two arrivals using waveform cross-correlation requires an operation of Hilbert transform of the first arrival before the regular cross-correlating. We investigated the SmKS waveforms of deep earthquakes occurring in South America recorded by several large and dense seismic arrays in China, and measured the differential arrival times of the SmKS pairs. We found significant waveform distortion of the SmKS caused by postcritical refraction and reflection at the CMB. For example, the π/2 phase between SKKS and S3KS cannot be confirmed from some major arc records at ~170°. This waveform distortion can introduce significant bias to the measured differential times, leading to incorrect estimate of P-wave velocity of the outer core.
Media has not been submitted for this Presentation
Abstract: The primary operational goal of the University of Utah Seismograph Stations (UUSS) is the collection and preservation of ground motion data. In 2007 the State of Utah provided funding for UUSS to design and build out a hardened seismic network. Initially, we took four steps toward realizing this goal: (1) a data center hot site was developed at the State of Utah backup facilities in Richfield, Utah, (2) data collection nodes were distributed across the State, (3) additional data nodes were built at radio relays on mountain tops, and (4) data feeds were established to the U. S. Geological Survey. This initial phase was targeted at removing the University of Utah as a single point of failure and providing multiple copies of seismic data for redundancy. Subsequently, data storage and computer redundancy have been increased at the mountain top relay sites and Raspberry Pi’s have been added to stations without adequate storage to prevent data loss in the case of telemetry failures. To make this system work, the SEEDlink protocol is used to backfill data, and a local Common Wave Buffer (CWB) and the IRIS DMC are populated with continuous data. The next phase of preserving data involves keeping up with the ever changing technological and security advances. Current work involves migrating all aspects of the UUSS network to IPv6 (the next generation of the Internet Protocol) due to the limitation on static IP numbers, and upgrading cell modems to support 4G and IPv6 to handle upgrades from our Internet Service Providers.
Media has not been submitted for this Presentation
Abstract: Funding agencies and network operators share an interest in understanding the performance benefits from investments in seismic stations and telemetry. The earthquake early warning (EEW) mission comes with additional requirements for coverage and resilience compared to conventional regional seismic monitoring. EEW alerts require greater spatial coverage, and network communications have to maintain that coverage after a damaging earthquake in order to alert on large aftershocks. Network reliability under these circumstances requires independent telemetry modes and spatial interleaving to reduce exposure single-point-of-failure vulnerabilities. To measure the contributions of individual stations and telemetry investments in the context of the EEW mission, we adapt the alert time calculator from Hotovec-Ellis et al. (SRL, 2017). EEW alert times to some central point are calculated using the current network assuming sources on a grid covering the network region. A candidate new station is then added and the calculation is repeated. The new station decreases time to alert in some area around it. We summarize its effect in a new metric with units of km^2-seconds, for the total area improved times the alert time improvement. The impact of station removal is computed the same way except that alerts are delayed. We can compute telemetry system impact metrics by removing or adding related stations as a group. The impact score can be modified by a proxy probability of earthquake occurrence in the affected area. We use these approaches to evaluate station and backhaul telemetry plans for the West Coast ShakeAlert EEW system. This analysis identified opportunities to improve the SCSN network cell service or internet dependence is significant. Station and hub improvements are being implemented. We have also used these results to draft new plans for West Coast network telemetry coverage.
Media has not been submitted for this Presentation
Abstract: The surge of using geophoned autonomous nodes for scientific projects is advancing passive imaging and monitoring techniques for scientific research, oil and gas projects, hydrology, civil engineering and new applications for short term dense seismic monitoring. The concept of a small, minimal configuration, low power, simple deployments without worry of environmental conditions can be applied to temporary broadband sensor system deployments using the next generation of direct bury broadband sensors. We explore some preliminary scientific deployment scenarios that can be used for broadband studies or as an example of a much smaller logistics next generation Earthscope array type station with little if any difference in noise performance and a potentially order of magnitude less cost. The concept of research grade latency can be applied for telemetry to recover quasi realtime data or very low cost system state of health.
Media has not been submitted for this Presentation
Abstract: The USGS/Caltech Southern California Seismic Network (SCSN) is a modern digital ground motion seismic network. It develops and maintains Earthquake Early Warning (ShakeAlert) data collection and delivery systems in southern California as well as real-time ShakeAlert algorithms. Here we present recent and ongoing innovations in telemetry, data quality control, and data analysis that keep the network running efficiently and provide timely high-quality streaming data. EEW systems must process seismic data within a fraction of a second of acquisition. The SCSN maintains a robust and resilient network of more than 350 seismic stations to achieve this goal. We have continued to improve the telemetry path diversity, minimize data transfer latency as well as developed new tools for latency monitoring and archiving. The latency data collection is done by an earthworm module that is user configured. It can collect latency channels under different names as mseed waveform data: L1Z (start of the packet), L2Z (middle of the packet) and L3Z (end of the packet) latency values. The latency data streams are archived with the same tools as the seismic data. We use PQLX for the data quality control because it enables monitoring of seismic signal performance for the entire network, and to identify anthropogenic noise sources and malfunctioning acquisition equipment. We have built a dynamic web-based display showing the summary noise analysis results. It sorts multiple channels separately and has mapping capabilities for an arbitrary number of the data points. This tool quickly identifies problematic sites and areas with elevated noise. It can also be used to report the worst N number of sites without analyzing all the data from the entire network.
Media has not been submitted for this Presentation
